Deoxyribonucleic acid (DNA) and Ribonucleic acid (RNA) are molecules that store and transmit genetic information and are present nearly all living organisms. The field of nucleic acid nanotechnology uses these molecules out of its biological context and employs it to build structures and then to connect their operation. Although DNA nanotechnology was originally developed to elucidate protein crystallization, recent developments in the field are testing the limits of its application towards a multitude of fields ranging from nanofabrication to computation. One of the important frontiers that remain to be addressed is the production of ‘active’ material that can interact with its environment and adapt to it intelligently, rivalling organelles present inside cells. This dissertation reports on the construction of responsive nucleic acid structures that can demonstrate autonomous function and capability to respond to physical and chemical inputs.

As the first example, we show how the self-assembly process of monomers made out of DNA strands can be triggered into activation by specific chemical inputs, in our case RNA molecules to build tubular structures. These ‘nanotubes’ can be temporally controlled by simple molecular programs, mimicking the architecture used by biological cells to direct their internal scaffolds. Our molecular programs use enzymes to produce or degrade RNA molecules embedded in the DNA nanotubes. Our results indicate that RNA can be used as a fuel for assembly, and that genetic circuits and enzymes can be an integral part in the operation of active nanostructures. This activatable self-assembly technique could be used to create a programmable synthetic version of filaments whose operation may mimic the cytoskeleton.

The second theme of this dissertation is the autonomous control of assembly and disassembly of nucleic acid nanotubes inside cell-sized environments. We used different designs to observe and control nucleation, polymerization and depolymerization steps in the self-assembly dynamics of encapsulated structures. The kinetics of growth and degradation of these encapsulated tubular structures were quantified using epi-fluorescence microscopy. We were also able to demonstrate the production of nucleotide assemblies with external stimuli, such as heat and light. These demonstrations can pave the way for designing and observing the functionality of responsive materials across the nanometer to micron size scales.

The ability to fully exploit redundant control capabilities available in aircraft is a critical feature in dealing with actuator failure scenarios. Managing input redundancy is traditionally addressed by means of Control Allocation (CA), where the aerodynamic control surfaces are typically considered only as moment-producing devices, thereby neglecting body forces, thus limiting a priori the possibilities offered by control reconfiguration. In the work presented in this thesis, we present a novel fault-tolerant control architecture for fixed-wing over-actuated and classical configuration small UAVs, where coupling between aerodynamic effectors and body forces are considered. The control architecture includes an adaptive loop for stable tracking of airspeed, flight path, and turn rate reference trajectories in nominal conditions (no actuator failures) in the presence of model parameter uncertainty. Fault tolerance is provided by a Dynamic Control Allocation (DCA) mechanism that automatically redistributes the control commands to the effectors as well as modifies certain reference trajectories to maintain the stability of the aircraft under multiple actuator failures. The particular design of the allocator makes fault identification algorithms unnecessary, simplifying the overall structure of the CA technique. In support of the theoretical development of this novel methodology, two case studies are presented. In the first case study, we consider flight simulations of a conceptual small over actuated UAV, which incorporates all the characteristics and the control architecture needed for the full exploitation of the flight control system herein developed. In the second test case, we consider real flight tests performed on a classical general aviation RC aircraft, which uses flaps as redundant control surfaces for roll control. Results are presented for different flights where ailerons are externally driven to move by 50% of the commanded deflection or not to move at all, simulating real ailerons failure conditions. Characterization of the environmental conditions is provided for each flight test in terms of normal accelerations, to quantify the level of perturbations faced by the flight controller. Simulation results are provided to demonstrate the performance and reliability of the control architecture, while flight test results are analyzed to confirm the robustness of the scheme against parameter uncertainty, actuation failure, and disturbance rejection.

Programmable, synthetic cells have applications in sensing and drug-delivery. Currently, development of synthetic cell components focuses on compartmentalization, and on developing the minimal machinery to carry out different cellular processes. In native cells, cytoskeletal filaments are a key structure for cell division, motility, and intra-cellular transport. Harnessing these filaments for use in synthetic systems is limited by the complexity of the dynamic behavior of the filaments. Alternatively, synthetic tile-based DNA nanotubes are comparable in length and stiffness to cytoskeletal filaments and can be engineered to demonstrate dynamic behavior using few components. To use DNA nanotubes as cytoskeletons in synthetic systems requires resilience to degrading enzymes found within cells, and the dynamic behavior must be automated and characterized in compartments.

Minimal cell systems execute tasks using transcription–translation (TXTL) machinery adapted from native cells. As other DNA nanotechnology degrades rapidly in vivo, I assayed the robustness of DNA nanotubes in an Escherichia coli cell-free TXTL system. TXTL recapitulates physiological conditions as well as strong linear DNA degradation through the RecBCD complex. I demonstrated that chemical modifications of the tiles composing DNA nanotubes and the addition of a Chi-site dsDNA, an inhibitor of the RecBCD complex, extend nanotube viability in TXTL for more than 24 hours. These complementary approaches are a first step towards engineering resilient DNA nanotubes for application in active environments.

To demonstrate autonomous control of assembly and disassembly processes of nanotubes in cell-sized environments, I implemented a DNA-RNA hybrid nanotube design inside of water-in-oil droplets. In this design, DNA tiles are activated by the presence of a trigger RNA molecule, which can be produced by and degraded by distinct enzymes. A pulse of nanotube assembly-disassembly occurs when both transcribing and degrading components are present with inactive tiles in droplets. Notably, the encapsulated system requires lower concentration of gene to trigger assembly than bulk solution. Varying the concentration of gene and degrading enzymes affects both the kinetics of assembly-disassembly and the morphology of the nanotubes. These methods can be employed to develop more complex dynamics and functionalities of DNA nanotubes as synthetic cytoskeletal filaments in minimal cells.

We derive a coarse-grained model that captures the temporal evolution of DNA nanotube length distribution during growth experiments. The model takes into account nucleation, polymerization, joining, and fragmentation processes in the nanotube population. The continuous length distribution is segmented, and the time evolution of the nanotube concentration in each length bin is modeled using ordinary differential equations. The binning choice determines the level of coarse graining. This model can handle time varying concentration of tiles, and we foresee that it will be useful to model dynamic behaviors in other types of biomolecular polymers.

We design a new biomolecular circuit with the potential for bistable behavior. We study this candidate toggle switch and find sufficient conditions on key parameters that guarantee bistability. The circuit structure is based on a positive feedback loop created by the mutual repression of two synthetic genes. Repression is generated not by direct inactivation of the promoter region, but by inactivation of the enzymes performing transcription. Inactivation is experimentally possible by designing the RNA outputs of the two genes to be inhibitory aptamers for the enzymes.

In a synthetic biological network it may often be desirable to maximize or minimize parameters such as reaction rates, fluxes and total concentrations of reagents, while preserving a given dynamic behavior. We consider the problem of parameter optimization in biomolecular bistable circuits. We show that, under some assumptions often satisfied by bistable biological networks, it is possible to derive algebraic conditions on the parameters that determine when bistability occurs. These (global) algebraic conditions can be included as nonlinear constraints in a parameter optimization problem. We consider two simple case studies for which we derive bistability conditions using Sturm's theorem, and optimize their nominal parameters to improve switching speed and robustness to a subset of uncertain parameters.

Biological systems reconfigure their shape in response to external stimuli at the level of single cells, tissues, and organs, for a variety of purposes such as growth, development, and self-repair. Cell shape reconfiguration is accomplished by directing the spatial organization of molecular materials (for example, cytoskeletal proteins) through molecular circuits which sense, process, and transmit information. Embedding a similar architecture in a synthetic material may greatly advance our ability to build responsive materials which can grow, reconfigure, and self-repair. DNA and RNA are programmable biological polymers which have been used to rationally build sensors and circuits, and a variety of nanostructures. The integration of nucleic acid circuits and structures promises to yield a new class of complex, reconfigurable materials. We aim at directing assembly and disassembly of DNA nanostructures with dynamic DNA inputs and circuits, mimicking the organization of dynamic cellular materials.

We focus on DNA nanotubes, and design pathways to control growth and breakage of tubes assembled from double crossover tiles. Our tiles include a toehold domain enabling strand invasion at the sticky ends of tiles. Invasion rapidly weakens the self-assembled structure, causing the nanotubes to collapse into smaller fragments. Removal of DNA species used for invasion, through second layer of strand displacement (denoted as anti-invasion), allows the nanotubes to reassemble isothermally. Our design thus makes it possible to control growth reversibly. We demonstrate this directed assembly process through a variety of experimental assays including optical microscopy and time-lapse movies, atomic force microscopy, and gel electrophoresis. We also characterize the influence of experimental design parameters such as toehold orientation (inside vs. outside of tube) and length, buffer conditions, and temperature, showing the overall tunability of breakage and reassembly dynamics.

Here, we bridge dynamic and structural DNA nanotechnology and demonstrate the programmable, dynamic control of self-assembly of DNA nanotubes, a well-known class of micron-sized DNA nanostructures. Nanotube assembly and disassembly is controlled with nucleic acid inputs and with transcriptional systems including an autonomous molecular oscillator.

Nucleic acid nanotechnology proposes many approaches to construct self-assembled structures using RNA and DNA. These structures have several applications such as scaffolding, sensing, and drug delivery. Although RNA and DNA are similar molecules, they also have unique chemical and structural properties that must be considered when using these molecules to assembly structures. RNA is a natural multifunctional polymer, and has an essential role in complex pathways and structures within the cellular environment. Due to its rich and diverse biological functions, synthetic self-assembling RNA nanostructures are emerging as a powerful tool with potential applications in drug delivery and intracellular control.

Here we demonstrate the formation of RNA nanostructures of varying size, complexity, and biological functionality using methods adapted from DNA nanotechnology. To achieve this goal we imported established approaches in the field of DNA nanostructure and DNA circuit design and optimized assembly protocols to ensure high yield of assembly. We employed design principles and experimental methods to obtain synthetic programmable RNA structures with biological functionality through two primary objectives: (1) Design and characterization of RNA tiles that assemble into RNA nanostructures with predictable features and (2) The usefulness of our RNA assemblies in a biomedical application, in particular functionalization with RNA molecules such as small interfering RNA (siRNA) for targeted RNA interference to inhibit gene expression. Our results show that scalable RNA self-assembled structures can be obtained using purely with Watson-Crick interactions, without relying on conserved tertiary structure motifs typically used to build RNA structures. Lastly, results show that functionalization does not prevent nanostructure assembly and these structures are capable of silencing genes.

Synthetic biology promise to provide solutions to many challenges in energy, agriculture, and health by reprogramming cells to execute new tasks in the host organism. In order to do that, it requires (1) the understanding of the design principles that underlie complex dynamics in biology, (2) the development of computational tools that support the identification those principles and (3) the use of those principles and computational tools to guide the experimental implementation of novel biomolecular programs. The main motivation of this thesis is to describe my current progress and future plans to expand (1)-(3).

We incorporate a new design principle, known as ultrasensitivity response, to de- sign robust biomolecular dynamical system. We show that molecular titration in the context of feedback circuits enhance the emergence of oscillations and bistable behavior in the parameter space. We also propose and analyze a new molecular network, termed Brink motif, which exhibits an ultrasensitive input-output response similar to a zero-order ultrasensitive switch. We discuss the Brink motif in the context of robust feedback circuits as a suitable mechanism to build (1) reliable circuits, oscillatory and bistable dynamical behaviors, under parameters uncertainty, downstream load effects and shared resources and (2) robust closed loop controllers that overcome the limitation of unidirectional action controllers. Ultrasensitivity is achieved by combining molecular titration and an activation/deactivation cycle and requires fast titration and switching rates. Additionally, the response of the Brink motif has a precisely tunable threshold, which can be determined by an external input to the motif. We assess the robustness of feedback circuits with numerical simulations and mathematical analysis.